Repository navigation
Conversation
Republish the tested implementation with a sign-off matching the GitHub commit author. Source is unchanged from ab3a8c4. Signed-off-by: jwc1love <155615614+jwc1love@users.noreply.github.com> Signed-off-by: jwc1love <jwc20050401@gmail.com>
Republish all seven reset regression tests unchanged from tested commit ab3a8c4. Signed-off-by: jwc1love <155615614+jwc1love@users.noreply.github.com> Signed-off-by: jwc1love <jwc20050401@gmail.com>
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
|
Important Draft PR not reviewedDraft PRs are not automatically reviewed by default.
To automatically review draft PRs, update your CodeRabbit configuration: reviews:
auto_review:
drafts: true
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
Hi @jwc1love. Thanks for your PR. I'm waiting for a tikv member to verify that this patch is reasonable to test. If it is, they should reply with Regular contributors should join the org to skip this step. Once the patch is verified, the new status will be reflected by the I understand the commands that are listed here. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. |
… L1 files Force at least three L1 SSTs, retain range tombstone fragments using an old snapshot, and compare both iterator directions with fresh scans and explicit full key/value expectations. Covers EOF, widening bounds and revisiting earlier files. The new regression passed in the canonical wrapper test target. Signed-off-by: jwc1love <155615614+jwc1love@users.noreply.github.com> Signed-off-by: jwc1love <jwc20050401@gmail.com>
Enable iterator reuse across bounded scan ranges for tikv/tikv#4239.
DBIterator::reset_bounds_and_seekreplaces owned bounds and performs a hard forward or reverse seek while retaining the snapshot and other read options.Nonemeans unsupported without mutation;Some(false)means a supported empty result. Titan, tailing, and missing/empty-bound configurations reject reset before mutation. Callers must honor column-family comparator and prefix constraints; TiKV rejects prefix-seek reuse. No native ABI or submodule change is required.Validation at
778cab5e4bcbfa3ebdc88fa7b38c8d97c0d0b43f: all 15 iterator cases passed (8 bounds-reset regressions and 7 existing tests), including a new case that asserts at least three L1 SSTs and retained range-tombstone fragments after compaction. It checks explicit key/value expectations and fresh iterators across forward/reverse file jumps, EOF, expanded bounds, empty ranges, reinsertion, and revisited ranges.Expanded validation uses the canonical
tests/tests.rsentry point through normal Cargo. The initial single-process suite aborts withSIGSEGV; all 143 cases were then executed independently: 140 passed and 3 crashed. These three also crash separately on unmodified base797c014576c9f01a854bd681711c617c5920c9c6with the same native cache:cases::test_rocksdb_options::test_block_based_optionscases::test_statistics::test_db_statisticscases::test_statistics::test_shared_db_statisticsThe root cause remains unresolved. Only these three cases were baseline-compared; this is not a successful single-process suite or a full baseline-suite comparison. Native RocksDB/Titan artifacts remained cached. Both repositories passed formatting checks. Downstream TiKV passes all six scanner integration cases, including real Region/Hybrid cache fallback, snapshot visibility, region bounds, locks, and statistics.
Verification report and separate follow-up evidence: https://github.com/jwc1love/tikv/blob/coprocessor/reuse-mvcc-scanner-4239/tests/benches/coprocessor_executors/scanner_reuse_4239.md
Supersedes #863 to correct its author/DCO identity. The original measured wrapper revision
ab3a8c…and republishedcc93046…have identical trees;778cab5…adds only the new regression test. Keep draft pending API/native-bound review and full CI; platform, release, and production validation remain outstanding.